Improving Reinforcement Learning Speed for Robot Control | IEEE Conference Publication | IEEE Xplore